The Management of Patients with Cerebrospinal Fluid Shunts

John D. Loeser, MD, Clifford J. Sells, MD, and David B. Shurtleff, MDSeattle, Washington

This article outlines the principles of management of the pa­tient with a cerebrospinal fluid shunt, emphasizing the clinical and laboratory methods of determining shunt malfunction or infection. Appropriate therapies for each complication are de­scribed.

Continuous patient evaluation and prompt treatment of complications are required to maximize the beneficial effects of a cerebrospinal fluid (CSF) shunting procedure for hydrocephalus. Not all patients can be closely followed by the neurosurgeon who inserted the original shunt; not all complications require the services of a neurosurgeon to initiate appropriate therapy. The purpose of this article is to outline a rational pro­gram for the assessment and management of a pa­tient with a CSF shunt and to emphasize key fac­tors in the determination of optimal continuing therapy.

From the Departments of Neurological Surgery and Pediatrics, University of Washington School of Medicine, Seattle, Washington. Drs. Loeser, Sells, and Shurtleff are Affiliates of the Child Development and Mental Retardation Center at the University of Washington. Requests for re­prints should be addressed to Dr. John D. Loeser, Depart­ment of Neurological Surgery, RI-20, University of Wash­ington, Seattle, WA 98195.

THE JOURNAL OF FAMILY PRACTICE, VOL. 6, NO. 2, 1978

Evaluation of the PatientThe general scheme for the evaluation of a pa­

tient with a CSF shunt is illustrated in Figure 1. This flow diagram will serve as the framework for the discussion of the management of a patient with a CSF shunt. Obviously, any evaluation must commence with a careful history and pertinent physical examination. A wide variety of CSF shunt problems have arisen that could have been readily solved if the patient or the patient’s parents could have described the type of shunt system implanted at another institution. As children with hydrocephalus live longer, it is becoming all too common to be unable to obtain the original opera­tive records. The name of the surgeon or hospital may not be recalled, the hospital records and x-rays may have been destroyed, or delays in ob­taining information may interfere with optimal care. It is strongly urged that copies of the opera­tive report describing the shunt system be given not only to the referring physician but also to the parents. The type of shunt is an important portion of the history and will influence both the diagnos­tic workup and some important aspects of the therapy of shunt malfunction.

Most patients with shunts are children; the his-

285

CEREBROSPINAL FLUID SHUNTS

Proximal portion of Shunt con serve OS ventriculostomy if obstruction is distal

Figure 1. Flow diagram illustrating steps in the evaluation of a patient w ith a cerebrospinal flu id shunt.

tory must include growth and developmental landmarks so that recent head growth and per­formance can be placed in the perspective of past achievements. An interval history since the time of shunt insertion or last revision may reveal periods of shunt dysfunction or suggestions of in­fection.

The physical examination should include a gen­eral assessment of growth and development, be havior, and the various organ systems; but must focus upon size and shape of the head, skull su­tures, scalp veins, fontanelles, and fundi. Cranial nerve, spinal cord, and long tract functions may reveal changes consistent with increased intra­cranial pressure. Signs of systemic infection, vas­cular or renal disease, or loculation of cerebrospi­nal fluid about the distal shunt must be sought;

286

only after a firm data base is established by the history and physical examination can one pru­dently proceed to evaluation of the shunt system. Merely depressing the flushing device does not suffice to determine shunt adequacy. Confirma­tion of shunt function is established by special dynamic studies.1,2

The first determination to be made is the pres­ence of “ overt shunt dysfunction” (Figure 1, rhomboid A). This includes such problems as fluid collection around the shunt, palpable disconnec­tion of shunt components, flushing device perma­nently depressed or incompressible, erosion of the shunt through the skin, or x-ray evidence of mal­position or disconnection. If overt shunt dysfunc­tion is observed, the shunt must be revised.

If overt shunt dysfunction is not established,

THE JOURNAL OF FAMILY PRACTICE, VOL. 6, NO. 2, 1978

CEREBROSPINAL FLUID SHUNTS

Table 1. Antibiotics of Choice in CSF Shunt Associated Infections

Organism AntibioticRoute of

Administration Dosage Frequency

S ta p h y lo co ccu s M e th ic il l in In tra v e n o u s 100 to 200 m g /k g /d a y q 4 to 6 h rIn tra v e n tr ic u la r 25 to 100 m g /d a y qd

S ta p h y lo co ccu s P e n ic illin In tra v e n o u s 25,000 to 300 ,000 q 2 to 6 h r

s u sce p tib le to a q u e o u s u n its /k g /d a y

aqueous p e n ic ill in c ry s ta ll in e

B -S trep to coccus P e n ic i l l in * In tra v e n o u s 25,000 to 300,000 q 2 to 6 h r

or a q u e o u s u n its /k g /d a y

S tre p to co ccu s v ir id a n s c ry s ta ll in e

E Coli G e n ta m ic in In tra m u s c u la r 5 to 7 m g /k g /d a y q 8 h r

o r In tra v e n tr ic u la r 2 to 4 m g /d a y qd

K lebs ie lla

P roteus

P seu do m on as G e n ta m ic in In tra m u s c u la r 5 to 7 m g /k g /d a y q 8 h r

a n d /o r In tra v e n tr ic u la r 2 to 4 m g /d a y qd

C a rb e n ic ill in In tra v e n o u s 400 to 500 m g /k g /d a y q 4 h r

^N e ith e r a q u e o u s p e n ic il l in n o r c a rb e n ic il l in s h o u ld be a d m in is te re d in tra v e n tr ic u la r ly o r in tra th e c a lly ;bo th are n e u ro to x ic w h e n in je c te d d ire c t ly in to th e c e n tra l n e rv o u s sys te m .

the physician must then ascertain from the history and physical examination and from tests of shunt function if the intracranial pressure is too high for the patient (Figure 1, rhomboid B). Elevated in­tracranial pressure is usually associated with rapid growth in skull circumference, bulging fontanelle, prominent scalp veins, and sutural diastasis. The child often manifests lethargy, nausea and vomit­ing, headache, retarded cognitive and motor de­velopment, and neurologic signs which may in­clude sixth-nerve palsies, coma, and papilledema. The clinical impression can be confirmed by measuring shunt transmission pressures, cere­brospinal fluid flow, and ventricular pressure as has been described.1'2 Symptomatic increased in­tracranial pressure mandates shunt revision.

THE JOURNAL OF FAMILY PRACTICE, VOL. 6, NO. 2, 1978

When either overt shunt dysfunction or increased intracranial pressure are diagnosed, it is impera­tive to determine if there is bacteriologic contami­nation of the cerebrospinal fluid and the shunt (Figure 1, rhomboid C). This is done by sampling the cerebrospinal fluid from the shunt system or ventricles for culture and sensitivity, cells, glu­cose, and protein.3 If there are no clinical signs of infection, the authors do not advocate delaying a necessary shunt revision to await the results of culture of the cerebrospinal fluid. Revising a shunt in the presence of CSF infection may result in continued and often more severe infection. When there are no clinical signs of infection, always sample the cerebrospinal fluid at the time of shunt revision. If a silent infection has been present, the

287

CEREBROSPINAL FLUID SHUNTS

operative specimen will facilitate early, appropri­ate antibiotic therapy.

TherapyShunt infection, and shunt dysfunction often

occur together; an aggressive management pro­gram is required to promptly handle both prob­lems. First, the patient must receive antibiotics appropriate for both the bacteria and the infection site. The route of administration is dependent upon the type of shunt and the locus of shunt obstruction. For example, drugs which do not penetrate the blood-brain barrier must be given directly into the ventricle. Table 1 lists the usual organisms associated with CSF shunt infection, and the dosage and route of administration of the commonly used antibiotics.4'8 The table, however, is only a general guide. In all instances, cultures and sensitivity tests should be performed. Mean inhibitory concentrations of antibiotics in the ven­tricular fluid drawn prior to each intraventricular dose can be used to determine appropriate an­tibiotic dosage and frequency of administration. The antibiotic concentration in the cerebrospinal fluid immediately prior to the next dose should be at least two times the mean inhibitory concentra­tion of the specific organism. The average shunted ventricular space is approximately 150 cc; individ­uals with enlarged ventricles should have doses of intraventricular medication increased pro­portionately. Clearance of antibiotics from the cerebrospinal fluid varies with each patient and should be evaluated carefully to avoid either inef­fectively low doses or the accumulation of toxic amounts; in most patients daily administration is usually adequate. If a perforable cerebrospinal fluid device with access to the ventricular CSF space has not been used, a Rickham or Ommaya reservoir should be placed to avoid cerebral man­tle destruction due to repeated ventricular taps. If the child has an unoperated myelo-meningocele, nonpenicillin agents may be introduced cautiously via the sac cerebrospinal fluid. Perforation of the sac should be through its margin and should trans­verse normal subcutaneous tissue, avoiding

288

hemangiomatous and nervous tissue as well as the membranes over the dorsal surface of the sac.

Intraventricular antibiotic therapy should con­tinue for 10 to 14 days. Systemic antibiotic therapy should continue for an additional three to four weeks.5,6

In addition to antibiotic therapy, the increased intracranial pressure must be immediately al­leviated. When the shunt obstruction is distal, some shunt systems permit tapping the flushing device. With other types of shunt, ventriculos­tomy is required. The brain does not tolerate the combination of infection and increased pressure; both must be vigorously treated. Thirdly, after the CSF pressure has been normalized and adequate antibiotic levels established in blood, cerebrospi­nal fluid, and tissues, the infected, dysfunctional shunt must be removed. After three to seven days on external ventriculostomy drainage and antibio­tics, an entirely new shunt system is inserted as far away from the old tract as possible. Antibiotics are continued for two to five weeks. One week after antibiotic cessation, appropriate cerebrospinal fluid and blood samples are taken to rule out con­tinued infection. The patient must be carefully ob­served to ascertain that a suppressed infection does not recur.

When shunt function is apparently adequate and intracranial pressure is not elevated, the man­agement is, of course, different. One must then consider the possibility of infection in a patient with a well-functioning shunt (Figure 1, rhomboid D). Usually, clinical assessment alone rules out infection; suspicious cases must be evaluated by sampling cerebrospinal fluid and blood. When gram-positive infection is documented, a three to six week course of the appropriate antibiotic is advocated. After a week without drugs, cerebro­spinal fluid and blood are again cultured. If the in­fection persists, it will be necessary to replace the shunt to clear the infection (Figure 1, rhomboid F). When gram-negative infection is present (Figure 1, rhomboid E), the authors believe that total shunt replacement and coverage with the appropriate antibiotics are required.

Most patients will have neither shunt dysfunc­tion nor infection. Their shunts will require no treatment at this evaluation interval. Periodically, the need for the shunt will be questioned (Figure 1, rhomboid G). Pressure and flow studies may seem to indicate that the shunt is not transmitting fluid.

THE JOURNAL OF FAMILY PRACTICE, VOL. 6, NO. 2, 1978

CEREBROSPINAL FLUID SHUNTS

Removal of this shunt is often a disaster; utmost caution is urged when contemplating shunt re­moval. If it is believed that the shunt can be safely removed, it is first prudent to ligate it in situ. After four to six weeks, if no problems have developed, the shunt may be removed. In fact, this is a very rare event in our institutions and this maneuver is not advocated.

The interval until the next evaluation is, of course, determined by the nature of the patient’s problem. The authors routinely see the patient one month after shunt revision. If the shunt system is functioning well, one year is a standard time for a return visit to our clinic. More frequent assess­ment by the primary care physician may also be required.

Antibiotic SelectionThe problem of selection of the appropriate an­

tibiotic for a shunt-related infection cannot be overemphasized. In order to plan rational therapy, the following facts must be known:1. Organism(s) causing infection2. Antibiotic sensitivities of these organism(s)3. Locus (or loci) of infection

a. ventricularb. meningealc. loculated intracraniald. within shunte. within skin incisionsf. intravascularg. intraperitoneal (or pleural, etc)

4. Distribution characteristics of antibiotics which could be used for each organism; also optimal routes of administration and dosages.

5. Type of shunt system.The most common errors involve failure to give

a drug which gets to the site of infection by the route of administration; ie, gentamicin must be placed directly into the ventricle to achieve ade­quate levels in most ventricular systems. This is readily accomplished with a nonvalved flushing device (such as the Foltz To-and-Fro Flusher), which is one reason the authors prefer this type of shunt system. If one cannot backfill the ventricles

THE JOURNAL OF FAMILY PRACTICE, VOL. 6, NO. 2, 1978

from the flusher, a ventriculostomy is required for intraventricular instillation. In contrast, chloram­phenicol administered systemically achieves ex­cellent levels in all CSF spaces.

Preliminary data have already suggested that computerized axial tomography will revolutionize the assessment of hydrocephalic patients. Ven­tricular size can be accurately determined at fre­quent intervals with minimal radiation and nonin- vasive techniques that do not increase risk of in­fection. Whereas clinical assessment will remain the primary determinant of shunt function, the strategies for detailed testing will change. The concept of management of a patient with a CSF shunt which is proposed here will be applicable even with this technological advance.

AcknowledgementSupported in part by Training Grant #NS 05211 from

the National Institute of Neurological and Communicative Disorders and Stroke, and Grant #C 112 from the National Foundation - March of Dimes.

References1. Hayden PW, Rudd TG, Dizmang D, et al: Evaluation of

surgically treated hydrocephalus by radionuclide clear­ance studies of the cerebrospinal fluid shunt. Develop Med Child Neurol (suppl) 32:72, 1975

2. Harbert JL, Haddad D, McCullough DC: Quantitation of CSF shunt flow. J Nucl Med 14:405, 1973

3. Myers MG, Schoenbaum SC: Shunt fluid aspiration: An adjunct in the diagnosis of cerebrospinal fluid shunt infection. Am J Dis Child 129:220, 1975

4. McLaurin RL: Treatment of infected ventricular shunts. Child's Brain 1:306, 1975

5. Shurtleff DB, Foltz EL, Weeks RD, et al: Therapy of staphylococcus epidermidis: Infections associated with cerebrospinal fluid shunts. Pediatrics 53:55, 1974

6. Sells CJ, Shurtleff DB, Loeser JD: Gram-negative cerebrospinal fluid shunt-associated infections. Pediatrics 59:613, 1977

7. Kramer PW, Griffith RS, Campbell RL: Antibiotic penetration of the brain. J Neurosurg 31:295, 1969

8. Schoenbaum SC, Gardner P, Shillito J: Infections of the cerebrospinal fluid shunts: Epidemiology, clinical manifestations, and therapy. J Infect Dis 131:54, 1975

289